61 hours of unplanned downtime per year—that's the hidden cost of chaining seven 99.9% reliable components together, and exactly the problem HP's new Tailscale integration in its Remote System Controller is designed to eliminate.
The Dependency Chain That Kills Redundancy
Out-of-band management exists for the moments when production access fails: OS crashes, network partitions, power events. The irony is that traditional recovery paths rely on the same fragile infrastructure they're supposed to bypass—centralized VPN appliances, dedicated management networks, jump hosts, public IP exposure, complex firewall rules, vendor relay services.
Chaining even reliable components kills availability fast. Seven components at 99.9% each yields 99.3% aggregate uptime. That's roughly 61 hours of downtime per year for the management plane you need most during an outage. I've lived this. I spent years building infrastructure at large companies, and a datacenter power event taught me the lesson: the enterprise VPN we'd installed for recovery was itself offline during the outage. Only a forgotten backup VPN that quietly reconnected saved us from driving to multiple datacenters.
How HP Makes Every RSC Device a Mesh Node
Peter Seiler from HP contacted Tailscale after one customer wanted their entire fleet of Remote System Controllers to join an existing tailnet—without bolting on extra VPN infrastructure. HP built a proof of concept running Tailscale directly on the RSC platform, which is based on Ubuntu on NVIDIA Jetson hardware. The management controller itself becomes a Tailscale node.
The integration lives inside the RSC configuration interface. Administrators enable it from the embedded UI—disabled by default—and enroll devices using either interactive authentication flows or auth-key-based enrollment. No dedicated management network, no centralized VPN appliance, no complex firewall rules. As long as the RSC has internet connectivity, it's reachable through the tailnet.
The Operational Math That Matters
HP's existing Remote System Management (RSM) environment already requires customers to allow connectivity to multiple AWS IoT endpoints, TURN services, asset storage gateways, and HP service endpoints. Each firewall rule means change control, cross-team coordination, and a new failure point. For deployments across retail locations, factories, or edge sites, that overhead compounds fast.
Tailscale removes most of that surface. Outbound internet connectivity is the only requirement. No exposed IPs, no VPN gateways, no relay services to maintain. The management plane becomes as resilient as the mesh itself—which is designed to route around failures, not depend on a single centralized point.
By embedding Tailscale directly into out-of-band management hardware, HP shifts the recovery path from a brittle chain to a distributed mesh. Next time the power flickers, your management interface might be the one thing that still works.
Source: Redundancy only matters if you can reach it
Domain: tailscale.com
Comments load interactively on the live page.